Hands can be a vector for transmitting pathogenic microorganisms to foodstuffs and drinks, and to the mouths of susceptible hosts. Hand washing is the primary barrier to prevent transmission of enteric pathogens via cross-contamination from infected persons. Conventional hand washing involves the use of water, soap, and friction to remove dirt and microorganisms. The availability of hand sanitizing products for use when water and soap are unavailable has increased in recent years. The aim of this systematic review was to collate scientific information on the efficacy of hand sanitizers compared with washing hands with soap and water for the removal of foodborne pathogens from the hands of food handlers. An extensive literature search was carried out using three electronic databases: Web of Science, Scopus, and PubMed. Twenty-eight scientific publications were ultimately included in the review. Analysis of this literature revealed various limitations in the scientific information owing to the absence of a standardized protocol for evaluating the efficacy of hand products and variation in experimental conditions. However, despite conflicting results, scientific evidence seems to support the historical skepticism about the use of waterless hand sanitizers in food preparation settings. Water and soap appear to be more effective than waterless products for removal of soil and microorganisms from hands. Alcohol-based products achieve rapid and effective inactivation of various bacteria, but their efficacy is generally lower against nonenveloped viruses. The presence of food debris significantly affects the microbial inactivation rate of hand sanitizers.
Foodborne diseases caused by consumption of contaminated food and beverages are considered some of the most common human diseases around the world (41). Norovirus, nontyphoidal Salmonella, Listeria monocytogenes, Clostridium perfringens, Campylobacter spp., and Toxoplasma gondii are the foodborne pathogens most commonly reported in the United States, causing 9.4 million episodes of foodborne illness, 55,961 hospitalizations, and 1,351 deaths per year (49). In the United Kingdom, the Food Standards Agency (23) estimated that more than 500,000 cases of food poisoning occur each year. These cases are caused by infections with Campylobacter spp., which is responsible for about 280,000 cases each year, C. perfringens, with about 80,000 cases, norovirus, with about 74,000 cases, and Salmonella, with the highest number of hospitalizations, about 2,500 each year. More than 320,000 cases of foodborne zoonotic disease are annually reported in the European Union. The most common microorganisms causing foodborne diseases in this region are Campylobacter spp., Salmonella, and viruses such as hepatitis A virus (HAV) and norovirus (17). Among 31 microorganisms causing foodborne diseases, five foodborne pathogens, known as the “top five,” have been identified by food safety experts as highly infective agents that can easily be transmitted by infected food handlers and can cause severe illness. These top five foodborne pathogens are norovirus, Salmonella Typhi (typhoid-like fever), Escherichia coli O157:H7 or other enterohemorrhagic and Shiga toxin–producing E. coli strains, Shigella spp., and HAV (58). Greig et al. (26) reviewed 816 reports of foodborne outbreaks from the United States, Canada, Europe, and Australia and identified 14 agents responsible for most of outbreaks in which food workers were implicated. The 14 main agents included norovirus (or probable norovirus), Salmonella enterica, HAV, Staphylococcus aureus, Shigella spp., Streptococcus Lancefield A and G, and parasites such as Cyclospora, Giardia, and Cryptosporidium.
Pathogenic microorganisms in food can originate in the food itself or its source, such as the growing, harvesting, or processing environment, or can be introduced into the food through cross-contamination and infected food handlers. In industrialized countries, infected food handlers have been identified as an important cause of foodborne illness (4, 27, 29). Up to one-third of outbreaks in Ireland (4) and 12% of outbreaks in the United Kingdom (19) are estimated to be caused by infected employees. In another study of foodborne illness outbreaks in restaurants in the United States, food handling by infected workers was identified as the main factor contributing to around two-thirds (65%) of foodborne illness outbreaks (29). Food service facilities, including restaurants and catered events, are the settings where most food worker associated–outbreaks originate (52), and contact with bare hands and failure to properly wash hands were the most frequently reported factors contributing to outbreaks (53). Thus, good personal hygiene and safe food handling practices are essential for preventing foodborne illnesses.
Hand hygiene through hand washing is the most important practice for preventing the spread of pathogens (6). Hand washing with water and soap is generally considered the “gold standard” method for removing dirt and transient microorganisms from hands. Plain soaps have minimal or no antimicrobial activity against bacteria and viruses, but their surfactant action combined with friction and final rinsing under water can effectively remove dirt, soil, and microbes from the outer layer of hand skin (35, 61). Over the past two decades, increasing interest has been focused on the use of hand cleansing products with antimicrobial activity, such as antimicrobial soaps or instant hand sanitizers, including both alcohol-based and alcohol-free preparations.
Antimicrobial soaps are preparations containing both a detergent and antiseptics or disinfectants with antibacterial activity, such as triclosan, chlorhexidine gluconate (CHG), and para-chloro-meta-xylenol (PCMX). Antimicrobial soaps are effective against gram-positive microorganisms, have moderate activity against viruses and tubercle bacilli, but are less effective against gram-negative microorganisms (30, 35).
Alcohol-based hand sanitizers and alcohol-based hand rubs (ABHRs) are instant hand hygiene products; their antimicrobial activity is due to the ability of alcohol to denature protein. These products usually contain 60 to 95% alcohol plus a thickening agent or humectant such as polyacrylic acid, glycerin, or propylene glycol to decrease the drying effect of the alcohol. ABHRs have documented microbiological activity against bacteria (21, 47), fungi, and some enveloped viruses such as human immunodeficiency virus, herpesvirus, adenovirus, and influenza and parainflu-enza viruses (20). Lower efficacy against nonenveloped (“naked”) viruses has been reported, and the level of inactivation seems to differ depending on the viruses tested, alcohol type, alcohol concentration, and time of exposure (12, 20, 21, 25, 28, 45, 46, 48).
Another group of instant hand products, the alcohol-free hand sanitizers such as compounds based on povidone-iodine, triclosan, or quaternary ammonium, has also attracted growing interest. Despite being historically recognized as less effective than ABHRs, more recent formulations prepared with benzalkonium chloride (BZK) have many advantages over ABHRs, including residual antimicrobial activity after use, less drying effect on hand skin, and stable efficacy after repeated use (13).
Use of waterless hand sanitizers as an alternative to conventional hand washing has long been debated. Despite some potential advantages over conventional water and soap (quicker and easier usage), instant hand products are generally considered to more effectively meet needs in hospital and health care settings rather than food preparation settings. ABHRs containing 60 to 95% alcohol are recommended as an alternative to hand washing in hospital and health care settings when hands are not visibly soiled (5). In contrast, use of these alternatives has been not been recommended in food establishments because of the inability of these products to remove fat and food debris from soiled hands (57). To date, little research has been conducted to examine the efficacy of hand disinfectants against transient microorganisms normally occurring on food workers' hands during food preparation. The present systematic review was conducted to examine the performance of various hand hygiene products against foodborne pathogens in food preparation settings.
MATERIALS AND METHODS
An extensive literature review was conducted in November 2014 using the electronic databases Web of Science, Scopus, and PubMed. The search was limited to articles published in English from 1990 to 2014. Search terms used were “efficacy of hand washing,” “efficacy of hand sanitizers,” “evaluation of hand sanitisers,” and “effect of hand hygiene products.”
Three preliminary criteria were adopted to select journal articles. Only articles that described levels of inactivation of foodborne pathogens (the actual pathogens, not surrogates), included a research approach with quantitative outcomes, and described studies undertaken in industrialized countries were included in this study. In contrast, all book chapters, studies carried out on microorganisms not involved in foodborne illness, and studies involving inactivation of foodborne microorganisms from raw food or food contact surfaces were excluded before analysis, based on the title and the abstract.
Once preliminary results matching search terms were obtained, data were extracted in three steps: duplicate articles were identified and removed, remaining titles and abstracts were screened for eligibility against inclusion criteria, and full text articles were retrieved and assessed in terms of their study design and scientific approach. All articles identified were then critically reviewed and included as appropriate to provide an overview of the topic.
Of the 2,108 records originally matching the search terms, 38 journal article abstracts were preliminarily screened for eligibility after duplicates were removed. Subsequent analysis of the full text of these articles permitted selection of the 28 articles included in this review (Table 1). Among the selected studies testing hand washing products against foodborne pathogens, 10 provided information on norovirus, 3 on HAV, 2 on L. monocytogenes, 14 on E. coli, 8 on S. aureus, and 1 on Salmonella. No scientific information was found for other pathogenic bacteria such as Campylobacter spp. and Bacillus cereus.
In addition to conventional water and soap or water only, products more generally tested against pathogenic bacteria and viruses included antibacterial liquid soaps, alcohol-based hand sanitizers, and non–alcohol-based sanitizers such as those containing triclosan, CHG, povidone-iodine, and quaternary ammonium, i.e., BZK or benzethonium chloride (BZT), 5-pyrrolidone-2-carboxylic acid (PCA), and copper sulfate pentahydrate (CS). Hand washing practices considered included use of soap and nailbrush (36), wash-sanitize consisting of using hand sanitizers after hand washing with water and soap (7, 16, 43), and a new trademarked hand hygiene regime SaniTwice (James Mann, Handwashing for Life, Libertyville, IL) consisting of a two steps: application of an excess of alcohol-based sanitizer with hand rubbing, wiping hands with a paper towel, and final application of alcohol-based sanitizer (15).
The relative efficacy of products was generally tested in vitro, ex vivo, and/or in vivo. Most of the in vitro studies involved experiments carried out using a suspension assay consisting of a standardized quantity of the target microorganism treated with increasing concentrations of the test product, with the aim of estimating the inactivation rate for each product (1, 10, 14, 20, 21, 24, 25, 42, 50, 51). In one in vitro study, inactivation rates of tested products were evaluated on latex gloves immersed in a solution of phosphate-buffered saline (PBS) or crab cooking water artificially contaminated with L. monocytogenes at 5 log CFU/ml (40). Ex vivo tests included experiments carried out on skin from a freshly killed pig (the pig skin method). The skin was treated with sanitizing products and then artificially contaminated with challenge microorganisms to test residual activity of tested products after use (9, 24, 31, 50). In vivo studies involved experiments carried out with selected human volunteers to estimate the efficacy of each tested product to remove or inactivate target microorganisms from artificially contaminated whole hands, finger pads, or gloves. The vast majority of published in vivo studies were carried out on hands or finger pads artificially contaminated with pure cultures of bacteria or viruses without the presence of food components or organic material (9, 14, 22, 25, 33, 34, 37, 39, 43, 51). In seven studies, the efficacy of hand washing products was evaluated in a food preparation setting with naturally and artificially soiled hands or gloves (7, 8, 15, 16, 36, 40, 44). In three studies, inactivation rates of products was evaluated on hands contaminated with virus suspensions prepared with other organic loads such as fetal bovine serum or feces (16, 32, 36). Other factors pertaining to food preparation settings such as hygiene of nails (36) and wearing rings when handling food have also been considered (60). A summary of the experimental conditions applied and main findings from in vitro, ex vivo. and in vivo evaluations in all studies included in this review is provided in Table 2. Information relating to specific pathogens is summarized below.
Because human norovirus (HuNoV) cannot be routinely cultured in vitro, determining the effectiveness of sanitizers and disinfectants against HuNoV is difficult. Methodologies used to estimate the level of virus reduction include the use of reverse transcription quantitative real-time PCR (RT-qPCR) to quantify the number of RNA copies of HuNoV extracted and purified from tested samples (37, 38, 42) and the use of cultivable surrogates such as feline calicivirus (FCV) and murine norovirus (MNV). Norovirus surrogates were generally tested alone as an alternative to HuNoV (9, 16, 25, 32, 34, 36, 51) or in parallel with HuNoV (42).
Liu et al. (38) compared the efficacy of an antibacterial soap, alcohol-based sanitizer containing 62% ethyl alcohol, and water rinsing for the removal of HuNoV from artificially contaminated finger pads. Ethanol-based hand sanitizer was the least effective hand product tested (0.34 ± 0.22-log reduction). The greatest reduction was observed for water rinse only (1.38 ± 0.49-log reduction) and antibacterial soap (1.1 ± 0.49-log reduction). In a separate study, Liu et al. (37) tested various commercially available hand hygiene products containing 62 to 95% alcohol against multiple HuNoV strains on finger pads. The results revealed a wide range of efficacy (0.10- to 3.74-log reduction), depending on the product and strain tested. The highest level of RNA reduction was achieved by a 70% ethanol gel containing additional ingredients that seem to potentiate the virucidal activity of the alcohol. A limitation of the study reported by the authors was the presence in the test products of PCR inhibitors that may have affected amplification and led to an overestimate of virus reduction.
In eight studies, the efficacy of hand sanitizers was evaluated against FCV and MNV. Experimental methods used to estimate virus inactivation included a virus-specific cytopathic effect test consisting of culturing posttreatment samples on a serial dilution of permissive host cells (9, 16, 25, 32, 34, 36, 51) and a plaque assay test in parallel with a TaqMan RT-qPCR assay (42). Park et al. (42) evaluated in vitro virucidal efficacy of seven hand sanitizers containing ethanol, triclosan, and chlorhexidine against both norovirus surrogates (i.e., FCV and MNV) and HuNoV. None of the products achieved significant RNA reduction when tested against HuNoV, whereas results for the norovirus surrogates differed between with the plaque assay and the RT-qPCR assay. A general lack of correlation between the two detection methods and different levels of inactivation of FCV or MNV were generally observed. Only a 72% alcohol (pH 2.9) ABHR reduced the infectivity of both FCV and MNV (3.4- and 2.6-log reductions, respectively) based on the plaque assay test, whereas no correlation was found between reduced infectivity and RNA reduction measured by the RT-qPCR assay. Conflicting results were also reported in two studies of the in vitro and in vivo efficacy of hand products against FCV and MNV. Gehrke et al. (25) tested three types of alcohol: ethanol, 1-propanol, and 2-propanol. In vitro experiments revealed that 50 and 70% 1-propanol was more effective (104-fold reduction) than ethanol and 2-propanol. In contrast, 70% ethanol achieved greater virus inactivation (3.78-log reduction) in vivo than did 1-propanol and 2-propanol (3.58- and 2.15-log reductions, respectively). Steinmann et al. (51) compared the virucidal activity of three ABHRs and three antimicrobial soaps. Results from suspension tests indicated a ≥5-log reduction of both FCV and MNV achieved by two of the three ABHRs tested, which was greater efficacy than that of the soaps tested (typically ≤3-log reduction). Conversely, the modified finger pad test carried out against MNV only indicated the superior antimicrobial activity of a povidone-iodine soap (4.62-log reduction) compared with the other ABHRs and soaps tested. Two studies were conducted to evaluate the in vivo efficacy of hand hygiene products against FCV only. Lages et al. (34) tested four ABHRs, three nonalcohol sanitizers, and two triclosan antimicrobial liquid soaps after exposure times of 30 s and 2 min. All products tested generally had limited efficacy; only one antimicrobial soap containing 10% povidone-iodine (≤2.67-log reduction) and one ABHR containing 95% ethanol (≤1.30-log reduction) achieved appreciable virus reduction compared with a water rinse tested in parallel. Czerwinski and Cozean (9) compared a novel hand sanitizer containing BZK, a 62% ABHR, an antibacterial liquid soap, and a water rinse. Apart from a promising level of inactivation achieved with the novel hand sanitizer (3.49-log reduction), generally <1-log virus reductions were obtained in all the other cases. Two studies were conducted to evaluate the efficacy of products on hands artificially contaminated with a fecal suspension of FCV. Kampf et al. (32) tested the efficacy of three ABHRs; greatest reduction (2.17 ± 1.06 log) in FCV was achieved with a hand sanitizer containing 95% alcohol. Lower concentrations of alcohol did not achieve a >1-log virus reduction. Lin et al. (36) compared six hand washing practices with contaminated natural and artificial nails. Use of soap and a nail brush achieved the highest log virus reduction (2.54 ± 0.57-log reduction) followed by hand washing with antibacterial soap (2.26 ± 0.42-log reduction) and then the combined use of soap and hand sanitizer (2.13 ± 0.93-log reduction). In contrast, the use of hand sanitizer alone had limited efficacy (0.86 ± 55-log reduction). Presence of long nails on treated hands significantly impacted the efficacy of all hand products tested. Edmonds et al. (16) compared four hand hygiene regimes on hands contaminated with a suspension of MNV prepared with 0.5% fetal bovine serum to mimic soiling with organic matter. Hand hygiene practices included an antimicrobial soap, a 70% alcohol gel, hand washing followed by hand sanitizing, and SaniTwice. Sanitizing with 70% alcohol gel was slightly more effective (2.6 ± 0.41-log reduction) than hand washing with antimicrobial soap (1.79 ± 0.29-log reduction). Greater viral reduction was achieved by SaniTwice (4.04 ± 0.33-log reduction) and by the combination of conventional hand washing and sanitizing (3.19 ± 0.31-log reduction).
Little published information is available about the relative effectiveness of hand washing products against HAV. Only three studies such studies were retrieved for this review (20, 21, 39). Fendler et al. (21) and Fendler and Groziak (20) found limited in vitro efficacy of a commercially available alcohol-based hand sanitizer containing 62% alcohol and emollients against HAV. After 30 s of exposure, 1.75-log (21) and 1.25-log (20) reductions were obtained, corresponding to 94.37 and 94.4%, respectively, reductions in the original inoculum.
Mbithi et al. (39) evaluated elimination rates obtained with 10 products on whole hands or finger pads artificially contaminated with a mixture of viruses and feces. Formulations tested included a nonmedicated soap, five ethanol-based hand sanitizers, and four antibacterial liquid soaps; tap water without soap was used as the control. None of the tested products resulted in 99.9% inactivation, which is generally desired. Inactivation rates observed for the whole hand and finger pad methods were 79 to 94%. One antibacterial soap and the nonmedicated soap attained a higher level of virus reduction (≤94.56% ± 5.75% and ≤91.39% ± 2.65%, respectively) than did alcohol-based hand sanitizers (≤90.67% ± 2.08%) and tap water (≤81.57% ± 4.5%). Residual infectivity, estimated as the mean number of PFU through a plaque assay test, was 0 to 0.64 PFU for ABHRs, 0.63 to 1.74 PFU for antimicrobial soaps, 1.57 PFU for plain soap, and 3.88 PFU for tap water. No published information was found concerning the efficacy of hand washing and hand sanitizers against HAV on hands soiled with food components.
Only two articles describing in vitro and in vivo efficacy of sanitizing products against L. monocytogenes were found (21, 40). Fendler et al. (21) reported a >5-log reduction of L. monocytogenes in vitro with a commercially available hand sanitizer containing 62% alcohol in a 30-s timed exposure kill test. McCarthy (40) compared the in vivo efficacy of one hand sanitizer and five disinfectants (two with chloride, one with iodine, one with peroxide, and one with quaternary ammonium) on contaminated latex gloves. The impact of the organic compounds on inactivation rates achieved with the tested products was estimated through immersion of gloves in both sterile phosphate-buffered saline (PBS) and crab cooking water artificially contaminated with L. monocytogenes at 5 log CFU/ml. Of the products tested, only the peroxide-based product achieved a 5-log reduction of attached L. monocytogenes on both soiled and nonsoiled contaminated gloves. The two chloride-based and the quaternary ammonium–based products achieved a 5-log reduction on gloves contaminated with PBS suspensions of L. monocytogenes (i.e., no food residues), but efficacy was lower (≤1- to 2-log reductions) in the presence of crab cooking water. The iodine-based sanitizer and alcohol-based instant hand sanitizer had lower efficacy in both experiments. No data about the efficacy of conventional hand washing for removing L. monocytogenes from gloves or hands was found in the literature.
S. aureus and E. coli.
Six studies included assessment of the in vitro and/or ex vivo efficacy of hand sanitizers against S. aureus and E. coli. Products included conventional ABHRs and new generation products containing a combination of active antimicrobial agents and other compounds such as thickening agents, emollients, and natural compounds. Fendler et al. (21) reported that a >5-log reduction was achieved by a 62% alcohol–based sanitizer against methicillin-resistant and vancomycin-tolerant S. aureus, nonpathogenic E. coli, and E. coli O157:H7. High in vitro inactivation rates were also reported by Biagi et al. (1), Czerwinski et al. (10), Gaonkar et al. (24), Kaiser et al. (31), and Shintre et al. (50). Biagi et al. (1) tested the in vitro efficacy of a new combination of two natural compounds, PCA and CS. The combination of PCA and CS was more effective than 70% ethanol and 60% isopropanol used alone. Czerwinski et al. (10) tested the efficacy of a novel alcohol-based antiseptic and a novel water-based antiseptic lotion prepared with a synergistic combination of ingredients centered on BZT. The novel water-based product had the same level of antimicrobial activity (99.9%) against E. coli and S. aureus strains as did the alcohol-based product. Gaonkar et al. (24) tested an ABHR prepared with an emollient (Octoxy) and other ingredients against E. coli and methicillin-resistant S. aureus. In vitro evaluations revealed a >7-log reduction of both E. coli and S. aureus, and ex vivo tests revealed greater antimicrobial activity and superior residual activity after use of the novel Octoxy formulation compared with the two other ABHRs applied in parallel as a control. Kaiser et al. (31) conducted an ex vivo comparison of a combination of a surgical scrub containing 4% CHG and ABHRs prepared with and without thickening agents against S. aureus. Hand sanitizers thickened with anionic polymers had a negative impact on persistent activity of CHG. In contrast, no negative effect was observed for ABHRs alone or those thickened with nonionic compounds. Shintre et al. (50) tested the synergistic effect of alcohol and quaternary ammonia in combination with moisturizers or essential oils in vitro and ex vivo. Synergistic combination of farnesol and BZT had longer activity (i.e., 20 to 35 min postapplication) against S. aureus and E. coli than did other hand sanitizers and chemicals compounds used alone.
The high level of bacterial inactivation generally observed in vitro does not necessarily reflect the actual capacity of products to remove transient microorganisms from the outer layers of skin on the hands. Incomplete effectiveness against target microorganisms from cleaned hands has been generally reported in all studies carried out on hands artificially contaminated with E. coli. Edmonds et al. (14) compared the efficacy of two novel 70% alcohol preparations (gel and foam), seven commercially available ABHRs, and two World Health Organization–recommended formulations containing 60 to 90% alcohol against one methicillin-resistant S. aureus strain. Results showed the superior efficacy of the novel gel and foam preparations after single and multiple uses compared with the other products. However, none of the products exceeded a 3-log reduction of the target microorganism. Fischler et al. (22) evaluated the effectiveness of two hand washing regimes for reducing transient bacteria after a single wash and subsequent potential transfer of bacteria to a ready-to-eat food. The antimicrobial soap achieved a higher level of inactivation (>3-log reduction) than did plain soap (≤2-log reduction) but failed to avoid the transfer of seeded bacteria to the ready-to-eat food item. Kampf et al. (33) reported limited efficacy of two ABHRs on hands artificially contaminated with E. coli. Bacterial inactivation achieved by two products containing 62% alcohol was only slightly better (≤3.5 ± 0.45-log and 3.58 ± 0.71-log reductions) than that achieved by rubbing hands under running water in parallel (2.39 ± 0.57-log reduction). Paulson et al. (43) examined the abilities of four hand washing regimes: plain soap, an antimicrobial soap, an alcohol-based hand sanitizer, and combined use of an antimicrobial soap and an ABHR (used after hand washing). All products used alone performed equally, but none exceeded a 2-log reduction. Higher efficacy (3.28-log reduction) was obtained with the combined use of hand washing and hand sanitizing.
Little published information about the efficacy of sanitizing products against Salmonella is available. Only one in vitro study was retrieved (21), in which a >5-log reduction of Salmonella Enteritidis and Salmonella Typhimurium was achieved by using a 62% ABHR.
Efficacy of hand products on hands soiled with food components.
Experimental conditions described in the literature to mimic food preparation settings include contamination of food workers' hands with natural soil encountered in the food service industry (7) and hands artificially inoculated with pure cultures of bacteria mixed with crab cooking water (40), chicken or beef broth (15, 16, 36), ground beef (7, 15, 16, 36), and dirt or cooking oil (44). The efficacy of the hand products was estimated based on enumeration of the microorganisms released from the treated hands or enumeration of bacteria remaining on the hands. Methods for enumerating released bacteria included evaluation of glove juice (15, 16, 60) or hand rinsate (8, 44). Both techniques consist of enumerating bacteria released from washed hands previously placed into a glove or a bag filled with sterile water or buffer. Conversely, enumeration of bacteria remaining on the hands after hand washing or hand sanitizing is usually estimated through image analysis or by pressing the palms of washed hand onto the surface of an agar plate (7).
Four studies included an evaluation of the efficacy of hand hygiene products on soiled hands. Courtenay et al. (8) compared the efficacy of three hand washing regimes (rinsing with warm water, rinsing with cold water, and washing with water and soap) for hands and gloves contaminated with E. coli and ground beef. Water and soap removed more bacteria than did the other hand hygiene regimes, and bacterial removal was higher from hands (99.98%) than from gloves (99.13%). The efficacy of four hand sanitizers containing 62% ethanol was also compared on clean hands contaminated with E. coli in broth at 106 log CFU/ml. The bacterial reduction achieved by the four hand sanitizers was 96.44 to 90.40% and was consistently lower than that obtained with water and soap. Charbonneau et al. (7) tested plain soap, a 70% alcohol-based hand sanitizer, and combined hand washing and alcohol-based hand sanitizer on hands naturally contaminated with raw chicken and ground beef. Higher efficacy was achieved with plain soap than with the other hand hygiene regimes. The limited efficacy of ABHRs on clean hands or hands soiled with dirt and oil was also reported by Pickering et al. (44). Bacterial reduction achieved did not exceed 2 log units of seeded E. coli (107 CFU/ml) in all cases. The efficacy of hand hygiene practices under moderate and heavy soil conditions was evaluated in two included studies (15, 16). Edmonds et al. (16) found superior efficacy for combined use of water and soap and hand sanitizing compared with water and soap or antimicrobial soap used alone; >5.0- and >4.6-log reductions of bacteria were achieved on moderately and heavily soiled hands, respectively. Edmonds et al. (15) tested the efficacy of the SaniTwice method with three 62 to 70% alcohol products compared with plain soap, an antibacterial soap, and a 70% alcohol gel used alone. SaniTwice with 70% alcohol foam had higher efficacy than did water and soap and other alcohol-based regimes; 4.61-and 3.92-log reductions of bacteria were achieved on moderately and heavily soiled hands, respectively. Heavy soil impacted the efficacy of all the practices tested (<1- to 2-log reductions).
Other considerations in relation to effective hand cleansing.
Only two included studies considered the efficacy of hand washing techniques for removal of bacteria or viruses from natural and artificial nails (36) or from hands with rings present (60). Wongworawat et al. (60) compared the efficacy of three hand sanitizers (one with povidone-iodine, one with water-aided alcohol, and one with alcohol and chlorhexidine) on hands with and without rings. The alcohol-chlorhexidine hand sanitizer was slightly more effective than the other products. No significant difference in the number of bacteria retrieved from cleansed hands with and without rings was generally observed. These results suggest that the presence of rings should not significantly impact the effectiveness of hand sanitizers.
Lin et al. (36) assessed the effectiveness of various cleansing products and hand practices for natural and artificial nails on hands inoculated with E. coli or FCV. Use of a nailbrush and soap achieved the highest removal of target microorganisms. However, the presence of long nails significantly impacted the efficacy of all regimes tested, suggesting that maintaining short fingernails is essential to reduce the risk of transmitting hazardous microorganisms when handling food.
Effective hand washing is extremely important to help prevent harmful microorganisms from spreading from people's hands to food. Contact with bare hands and failure to properly wash hands have been reported as the main risk factor contributing to foodborne disease caused by food handlers (53). European Union food safety legislation requires every person working in a food handling area to maintain a high standard of personal cleanliness and requires food business operators to provide an adequate number of wash basins suitably located and designed for cleaning hands (18). The Food Code 2009 (56), published by the U.S. Food and Drug Administration to standardize food safety and food hygiene procedures, states that the total time recommended for proper hand washing is at least 20 s, of which 10 to 15 s should be used for rubbing followed by rinsing under running warm water and then drying.
The presence of food components such as fat, oil, or other dirt is considered the main factor affecting inactivation rates of hand hygiene products against microorganisms on the hands of food workers (59). The levels of microbial contamination on hands of food workers have been reported as 2 to >5 log CFU per hand across various food settings, and the bacterial flora generally encountered on the hands of food workers is a mixture of Enterobacteriaceae and other mesophilic bacteria in the presence of fat and other soil (11). Various pathogens with very low infective doses (1 to 100 units), including viruses, parasites, and enteric bacteria, can be present in high numbers on contaminated hands (54). Pathogens carried by contaminated hands can be easily transferred to food and hand contact surfaces and can survive for long periods (54, 55). The ideal hand hygiene regime to be used in a food setting would ensure maximum removal of food components and food flora from cleaned hands to minimize the level of transferable microorganisms. Most of the hand disinfectants, including medicated soaps and instant hand sanitizers, have broader antimicrobial activity than do plain soaps but are generally considered to not properly meet the needs of food workers because these products are unable to remove food soil from cleansed hands (57).
In this systematic review, we evaluated the published scientific information available on the efficacy of conventional and improved hand hygiene products in relation to their use in food preparation settings. Analysis revealed the existence of conflicting reports about the efficacy of soaps and hand sanitizers against foodborne pathogens. No standardized method to estimate removal and inactivation rates of target pathogens is available, and differences in the experimental conditions (e.g., quantity of product used, duration of treatment, and type of food soil used) among studies makes comparison of results difficult. Hand washing with water and soap is generally reported to achieve effective removal of bacteria and soil from hands (7, 8, 15, 16) and gloves (8) and to be superior to use of other products when used with a nailbrush for removal of bacteria and viruses from fingernails (36). However, some microorganisms remain even after proper washing (7, 8, 15, 16, 36), suggesting that hand washing alone cannot ensure elimination of risk in relation to transmission of bacteria from hands to food. Conventional hand washing is more effective for contaminated hands than for gloves (8), suggesting that frequent changes of gloves rather than washing gloves when they become visibly soiled would more effectively minimize risk of bacterial contamination between food preparation steps.
The efficacy of antimicrobial soaps versus conventional plain soaps is also controversial, and these conflicting results have been reported in two other reviews (27, 61). Apart from one study (43), the evidence seems to indicate that antimicrobial or medicated soaps can achieve a slightly higher level of microbial inactivation on artificially contaminated hands without food residue present (22, 39), whereas their efficacy on soiled hands is similar to that of conventional soaps (16, 36).
Instant hand sanitizers have high and rapid in vitro efficacy against various target bacteria (10, 14, 21, 50), but their efficacy against naked viruses seems to be lower (20, 21, 34, 42) and differs depending on the virus, the type of alcohol, and the alcohol concentration (25, 42, 51). These findings are in general agreement with those of four other reviews (2, 3, 27, 61). Apart from some improved formulations (9, 28), instant hand sanitizers used in vivo do not usually exceed 2- to 3-log microbial reductions (14, 33, 37, 38, 43, 44), and their efficacy seems to be affected by the presence of food debris, as observed on both moderate (40) and heavily soiled (7, 15, 36) hands; only one study included in the present review reported similar rates of bacterial inactivation on both clean and soiled hands (44). Instant hand sanitizers used alone do not seem to be a reliable substitute for conventional hand washing in food establishments (7). In contrast, use of these sanitizers after hand washing with either antimicrobial or plain soap (i.e., wash-sanitize regimes) seems to be more effective than use of hand sanitizer or soaps alone (16, 43); bacterial inactivation significantly increased up to 4- or 5-log reductions on both moderately and heavily soiled hands (16).
Preliminary results reported for SaniTwice method are also encouraging (15). When the method was tested on hands moderately and heavily soiled with a mixture of food components and E. coli, adequate levels of bacterial control (~4-log reductions) were achieved. A similar level of inactivation has also been reported against MNV on artificially contaminated hands. These findings suggest that this hand hygiene regime could be used as an alternative to wash-sanitize when water and soap are not available. However, no published information about the efficacy of this hand hygiene regime against HuNoV or HAV on soiled hands seems to be available. For this reason, further studies are needed to evaluate the effectiveness of the SaniTwice method in different food settings and against different foodborne pathogens.
A new generation of alcohol-free lotions is attracting more interest (1, 24, 31). Evidence from in vitro and ex vivo studies indicate efficacy against target bacteria similar to that of alcohol-based products, with prolonged activity after application and potentially less skin irritation. However, very little is known about the efficacy of these alcohol-free products against viruses, and no information about their inactivation rates on soiled hands is currently available.
This material is based upon work supported by Safefood, from the Food Safety Promotion Board, under grant 06-2014.